Research On Classification De-noising And Cavity Repair Algorithm For 3D Laser Scanning Point Cloud Data

As a new type of non-contact and fast acquisition technology,three-dimensional laser scanning technology has been applied in many fields.How to improve the processing efficiency and precision of a large number of point cloud data is directly related to the application rate of the technology in various fields.Due to the influence of various systems and contingencies,the noise data in the point cloud data will not only increase the amount of cloud data,but will not only affect the precision and efficiency of the later modeling,there will be some noise data mixed in the point cloud data,these noise will not only increase the number of point cloud data,but also will be directly affect the accuracy and efficiency of late modeling if it is not processed in time.In addition,the noise points are usually divided into two kinds,the large-scale noise away from the main point cloud and the small scale noise mixed in the object model surface.At present,most of the algorithms are usually using the same approach for different types of point cloud data,although it can remove the noise,but it will therefore delete a lot of their own characteristics of the model and caused over smooth problems,what's more,it will lead to the distortion of the measured object after modeling.In this paper,a classification denoising algorithm for different types of point cloud noise data is proposed,it can realize the efficient denoising for point cloud data.The main of this paper is as follows:1.In the process of point cloud denoising,Based on the advantages of PCL processing for point cloud data and PCL integrated point cloud filter module to complete the classification denoising of point cloud data.Statistical Outlier Removal filter and Radius Outlier Removal filter is used to remove large scale noise,and then through the improved bilateral filter algorithm in the BilateralFilter class to realize the small-scale point cloud noise removal.Finally,a concrete engineering example is used to verify the feasibility of the classification denoising algorithm.2.When there is empty point cloud data in the process of denoising,a three-dimensional laser scanning point cloud data denoising algorithm based on RBF neural network is proposed.TThe RBF neural network is used to predict the cloud data which need to be processed,and the measured values of the point cloud are replaced by the network prediction value to realize the denoising of the point cloud data.The results show that the algorithm can effectively remove the noise data from the cloud surface,keep the characteristics,and fill the voids in the process of scanning and denoising.